1. Technical Field
This invention relates to the melting of titanium or titanium alloys in a plasma cold hearth furnace. More particularly, this invention relates to a plasma cold hearth melting method and apparatus for providing a titanium ingot of commercial quality. Specifically, the invention is a method and apparatus for optimizing melting using a combination of plasma torches and direct arc electrodes, each of which is extendable and retractable into the melting environment and moveable in a circular pivoting or side to side linear motion.
2. Background Information
For many decades, aircraft engines, naval watercraft hulls, high tech parts for machinery and other critical component users have used substantial amounts of titanium or titanium alloys or other high quality alloys in the engines, the hulls, and other critical areas or components. The quality, tolerances, reliability, purity, structural integrity and other factors of these parts are critical to the performance thereof, and as such have required very high quality, advanced materials such as ultra-pure titanium or titanium alloys.
For decades, titanium usage was only where critical to meet very high quality, tolerances, reliability, purity, structural integrity and other factors because of the high cost of the manufacturing process which was typically a vacuum arc re-melting (VAR) process. However, high density inclusions and hard alpha inclusions were still sometimes present presenting the risk of failure of the component—a risk that is to be avoided due to the nature of use of many titanium components such as in aircraft engines. High-density inclusions, also called HDIs, are particles of significantly higher density than titanium and are introduced through contamination of raw materials used for ingot production where these defects are commonly molybdenum, tantalum, tungsten, and tungsten carbide. Hard alpha defects are titanium particles or regions with high concentrations of the interstitial alpha stabilizers, such as nitrogen, oxygen, or carbon. Of these, the worst defects are usually high in nitrogen and generally result from titanium burning in the presence of oxygen such as atmospheric air during production. It is well known in the industry that the VAR process, even with the inclusion of premelt procedural requirements and post-production nondestructive test (NDT) inspections has proven unable to completely exclude hard alpha inclusions and has shown only a minimal capability for eliminating HDIs. Since both types of defects are difficult to detect, it is desirable to use an improved or different manufacturing process.
In more recent years, the addition of cold hearth or “skull” melting as an initial refining step in an alloy refining process has been extremely successful in eliminating the occurrence of HDI inclusions without the additional raw material inspection steps necessary in a VAR process. The cold hearth melting process has also shown promise in eliminating hard alpha inclusions. However, in many applications the plasma cold hearth-melting step is followed by a final VAR process since it provides known results. This is detrimental however as it risks reintroducing inclusions or impurities into the ingot. It is clear that a cold hearth melt only process would be more economical as a source for pure titanium than a VAR process or a hearth melting and VAR combination process.
The cold hearth melting processes currently being used incorporate either plasma or electron beam (EB) energy. It has been discovered that the cold hearth melt process is superior to VAR melting since the molten metal must continuously travel through a water cooled hearth before passing into the ingot mold. Specifically, separation of the melting and casting zones produces a more controlled molten metal residence time which leads to better elimination of inclusions by mechanisms such as dissolution and density separation.
However, additional improvements are needed to reach the ultimate potential that cold hearth melting using plasma or electron beam energy has to offer. Numerous issues still exist that result in a lack of optimization of the cold hearth melts process.
In electron beam cold hearth melting, a sophisticated and expensive “hard” vacuum (a vacuum at 10-6th millibars) system is still critical since electron beam energy guns will not operate reliably under any atmosphere other than a “hard” or “deep” vacuum. This vacuum also far exceeds the vapor pressure point of aluminum, which is often an element in titanium alloys. As a result evaporation of elemental aluminum results in potential alloy inconsistency and furnace interior sidewall contamination. Often sophisticated modeling and very thorough and costly scrap preparation are necessary due to the aluminum evaporation, as well as the addition of master alloys to make up for alloy evaporation losses. It is known that significant guesswork is often involved in making this process work.
In both plasma and electron beam cold hearth melting, many stirring and mixing inefficiencies exist. It is known that the more vigorous the stirring in a melting hearth the faster high melting point alloy additions go into solution, that a good homogeneous mixture requires enough stirring to reduce the potential for alloy segregation and that vigorous stirring insures against temperature variations in the melt hearth. It is also known that these temperature variations can make it difficult to reach a useful superheat.
The removal of high-density inclusions and hard alpha inclusions in a plasma and electron beam cold hearth melting process is also challenging. In operation, the residence time in the bath and a certain level of bath agitation resulting from the heat source are counted upon to “sink” the HDIs to the “mushy” zone at the bottom and “breakup” the LDIs to non-detectable levels. Experience has shown this to be an effective method of removing inclusions, however the process is certainly far from perfect and failure to remove the inclusions can be catastrophic.
Plasma and electron beam cold hearth melting are both continuous processes. From a practical standpoint, it is very difficult to sample the process as it occurs and therefore the results of the melt campaign are generally not known until the entire process is completed where product can be removed and physically sampled after cool-down. This has a number of associated drawbacks. First, it takes time before the plant knows whether the product is saleable. If the results are negative often the ingot is scrapped or must be cut up and re-melted again. Second, if the product can be salvaged it is usually downgraded and sold for less. Third, there are typically variations in chemistry throughout the product, which may be acceptable in an application but clearly point out the weakness in continuous operations of this nature. Even with good modeling capability the process is, at best, hit or miss. This is the primary reason most hearth melts require subsequent melting a second or third time in a conventional VAR furnace.
The continuous process also often does not yield a satisfactory surface finish. The result is the end user machining down the ingot prior to use. This is a large waste of resources—both in time and effort to machine the ingot, and in wasted titanium that is machined off into generally worthless titanium turnings or shavings.
It is thus very desirable to discover a method of re-using the inexpensive and readily available scrap or processed titanium turnings which have in the past been unusable in any quantity due to the high levels of surface oxygen contained therein as well as the potential and/or likelihood of molybdenum, tantalum, tungsten, and tungsten carbide contamination from machining with tool bits made of these materials.